• The Journal of the Acoustical Society of Korea
• /
• v.35 no.4
• /
• pp.303-309
• /
• 2016

This paper proposes a novel approach to the problem of audio signal clustering using a stacked autoencoder. The proposed stacked autoencoder learns an efficient representation for the input signal, enables clustering constituent signals with similar characteristics, and therefore the original sources can be separated based on the clustering results. STFT (Short-Time Fourier Transform) is performed to extract time-frequency spectrum, and rectangular windows at all the possible locations are used as input values to the autoencoder. The outputs at the middle, encoding layer, are used to cluster the rectangular windows and the original sources are separated by the Wiener filters derived from the clustering results. Source separation experiments were carried out in comparison to the conventional NMF (Non-negative Matrix Factorization), and the estimated sources by the proposed method well represent the characteristics of the orignal sources as shown in the time-frequency representation.

• The Journal of the Acoustical Society of Korea
• /
• v.27 no.5
• /
• pp.244-251
• /
• 2008

In this research, we propose an unsupervised debate indexing method using audio and visual information. The proposed method combines clustering results of speech by BIC and visual by distance function. The combination of audio-visual information reduces the problem of individual use of speech and visual information. Also, an effective content based analysis is possible. We have performed various experiments to evaluate the proposed method according to use of audio-visual information for five types of debate data. From experimental results, we found that the effect of audio-visual integration outperforms individual use of speech and visual information for debate indexing.

• The KIPS Transactions:PartB
• /
• v.19B no.1
• /
• pp.19-26
• /
• 2012

The rapid increase of information imposes new demands of content management. The purpose of automatic audio segmentation and classification is to meet the rising need for efficient content management. With this reason, this paper proposes a high-accuracy algorithm that segments audio signals and classifies them into different classes such as speech, music, silence, and environment sounds. The proposed algorithm utilizes support vector machine (SVM) to detect audio-cuts, which are boundaries between different kinds of sounds using the parameter sequence. We then extract feature vectors that are composed of statistical data and they are used as an input of fuzzy c-means (FCM) classifier to partition audio-segments into different classes. To evaluate segmentation and classification performance of the proposed SVM-FCM based algorithm, we consider precision and recall rates for segmentation and classification accuracy for classification. Furthermore, we compare the proposed algorithm with other methods including binary and FCM classifiers in terms of segmentation performance. Experimental results show that the proposed algorithm outperforms other methods in both precision and recall rates.

• International Journal of Computer Science & Network Security
• /
• v.21 no.8
• /
• pp.87-96
• /
• 2021

The demand for e-learning through video lectures is rapidly increasing due to its diverse advantages over the traditional learning methods. This led to massive volumes of web-based lecture videos. Indexing and retrieval of a lecture video or a lecture video topic has thus proved to be an exceptionally challenging problem. Many techniques listed by literature were either visual or audio based, but not both. Since the effects of both the visual and audio components are equally important for the content-based indexing and retrieval, the current work is focused on both these components. A framework for automatic topic-based indexing and search depending on the innate content of the lecture videos is presented. The text from the slides is extracted using the proposed Merged Bounding Box (MBB) text detector. The audio component text extraction is done using Google Speech Recognition (GSR) technology. This hybrid approach generates the indexing keywords from the merged transcripts of both the video and audio component extractors. The search within the indexed documents is optimized based on the Naïve Bayes (NB) Classification and K-Means Clustering models. This optimized search retrieves results by searching only the relevant document cluster in the predefined categories and not the whole lecture video corpus. The work is carried out on the dataset generated by assigning categories to the lecture video transcripts gathered from e-learning portals. The performance of search is assessed based on the accuracy and time taken. Further the improved accuracy of the proposed indexing technique is compared with the accepted chain indexing technique.

• Convergence Security Journal
• /
• v.23 no.4
• /
• pp.61-70
• /
• 2023

Clustering is an unsupervised learning method that involves grouping data based on features such as distance metrics, using data without known labels or ground truth values. This method has the advantage of being applicable to various types of data, including images, text, and audio, without the need for labeling. Traditional clustering techniques involve applying dimensionality reduction methods or extracting specific features to perform clustering. However, with the advancement of deep learning models, research on deep clustering techniques using techniques such as autoencoders and generative adversarial networks, which represent input data as latent vectors, has emerged. In this study, we propose a deep clustering technique based on deep learning. In this approach, we use an autoencoder to transform the input data into latent vectors, and then construct a vector space according to the cluster structure and perform k-means clustering. We conducted experiments using the MNIST and Fashion-MNIST datasets in the PyTorch machine learning library as the experimental environment. The model used is a convolutional neural network-based autoencoder model. The experimental results show an accuracy of 89.42% for MNIST and 56.64% for Fashion-MNIST when k is set to 10.

• Journal of Korea Multimedia Society
• /
• v.16 no.4
• /
• pp.453-463
• /
• 2013

In this paper, we present an effective method and a system for the music summarization which automatically extract the chorus portion of a piece of music. A music summary technology is very useful for browsing a song or generating a sample music for an online music service. To develop the solution, conventional automatic music summarization methods use a 2-dimensional similarity matrix, statistical models, or clustering techniques. But our proposed method extracts the music summary by calculating BER(Bit Error Rate) between audio fingerprint blocks which are extracted from a song. But we could directly use an enormous audio fingerprint database which was already saved for a music retrieval solution. This shows the possibility of developing a various of new algorithms and solutions using the audio fingerprint database. In addition, experiments show that the proposed method captures the chorus of a song more effectively than a conventional method.

• The KIPS Transactions:PartD
• /
• v.9D no.5
• /
• pp.779-788
• /
• 2002

There are many technically difficult issues in searching multimedia data such as image, video and audio because they are massive and more complex than simple text-based data. As a method of searching multimedia data, a similarity retrieval has been studied to retrieve automatically basic features of multimedia data and to make a search among data with retrieved features because exact match is not adaptable to a matrix of features of multimedia. In this paper, data clustering and its indexing are proposed as a speedy similarity-retrieval method of multimedia data. This approach clusters similar images on adjacent disk cylinders and then builds Indexes to access the clusters. To minimize the search cost, the hashing is adapted to index cluster. In addition, to reduce I/O time, the proposed searching takes just one I/O to look up the location of the cluster containing similar object and one sequential file I/O to read in this cluster. The proposed schema solves the problem of multi-dimension by using clustering and its indexing and has higher search efficiency than the content-based image retrieval that uses only clustering or indexing structure.

• The KIPS Transactions:PartC
• /
• v.9C no.1
• /
• pp.107-114
• /
• 2002

In these days, rapid advances of the computer system and the high speed network have made the multimedia services popularized among various applications and services in the internet. Internet live broadcasting, a part of multimedia services, makes it possible to provide not only existing broadcasting services including audio and video but also interactive communications which also expand application scopes by freeing from both temporal and spatial limitation. In the Paper, an interned Personal-live broadcasting server system is developed by allowing individual users to actively create or join live-broadcasting services with such basic multimedia devices as a PC camera and a sound card. As the number of broadcasters and participants increases, concurrent multiple channels are established and groups are to be expanded. The system should also guarantee High Availability (HA) for continuous services even in the presence of partial failure of the cluster. Furthermore, a transmission mode switching is supported to consider network environments in the user system.

• 한국HCI학회:학술대회논문집
• /
• 2006.02a
• /
• pp.400-405
• /
• 2006

In this paper, we present effective methods for music summarization which summarize music automatically. It could be used for sample music of on-line digital music provider or some music retrieval technology. When summarizing music, we use different two methods according to music length. First method is for finding sabi or chorus part of music which can be regarded as the most important part of music and the second method is for extracting several parts which are in different structure or have different mood in the music. Our proposed music summarization system is better than conventional system when structure of target music is explicit. The proposed method could generate just one important segment of music or several segments which have different mood in the music. Thus, this scheme will be effective for summarizing music in several applications such as online music streaming service and sample music for Tcommerce.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.56 no.5
• /
• pp.987-989
• /
• 2007

This paper describes a digital convergence platform(DCP) which is implemented based on the MPEG-21 multimedia framework. DCP is a newly proposed solution in this research for the convergence service of future home multimedia environment. DCP is a common platform designed to have the feature of reconfigurability, by means of S/W. which supports to run diverse digital multimedia services. A distributed peer to peer service and transaction model is also a new feature realized in DCP using the MPEG-21 multimedia framework. A prototype DCP is implemented to verify its functions of multimedia service and transactions. The developed DCPs are networked with IP clustering storage systems for the distributed service of multimedia. Successful streaming services of the MPEG-2/4 video and audio are verified with the implemented test-bed system of DCP.